Controlling inappropriate behavior in public online multiuser systems is challenging, particularly when these systems have millions of user accounts, such as popular multiplayer online games, e.g., League of Legends from Riot Games (www.leagueoflegends.com). These games are generally low cost to participate in and users can register anonymously. Moreover, these games can be very competitive, thereby enabling aggressive and undesirable behavior, such as cheating or inappropriate communications or any unacceptable behavior as determined by a player's peers. For example, during a game, an aggressive user may send text messages to other users with profanity, insults, racial slurs, and/or threats of violence. In a system with millions of users, this behavior may by exhibited by hundreds or thousands of participants if unchecked, and having the game administrator police every user and/or incident is highly burdensome and expensive. Moreover, it may be desirable to reward positive behavior or exceptional skill in such systems. Accordingly, an improved system for encouraging and discouraging certain behavior and game play would be desirable.